1. Switch to ATV/r-containing regimen  ATAZIP

2. Mallolas J, JAIDS 2009;51:29-36 ATAZIP ATAZIP Study: Switch LPV/r to ATV/r  Design  Endpoints –Primary: non inferiority in the proportion of patients with treatment failure at W48 (intent-to-treat analysis), lower limit of the 95% CI for the difference = -12.5%, 80% power –Treatment failure = virologic rebound (2 consecutive HIV-1 RNA ≥ 200 c/mL), lost to follow-up, withdrawn consent, discontinuation for any reason, progression to a new CDC event or death Switch to ATV/r 300/100 mg qd + continue NRTIs** Continue LPV/r 400 /100 mg bid + NRTIs** * Not more than 2 previous virologic failure on PI and if genotype performed < 5 mutations ** Switch in NRTI not counted as failure Randomisation 1 : 1 Open-label 265 patients HIV+ ≥ 18 years On LPV/r + 2 NRTIs ≥ 6 months HIV RNA 6 months* N = 127 N = 121 M24

3. ATV/r N = 121 LPV/r N = 127 Median age, years4243 Female20%21% History of AIDS diagnosis40%57% Hepatitis C co-infection36%27% CD4 cell count at baseline < 200 /mm 3 8%10% NRTIs: TDF + 3TC / TDF + ddI / ZDV + 3TC20% / 17% / 10%17% / 9% / 16% Previous PI failures ≥ 121%20% Previous PI mutations ≥ 1 ≥ 1 major 36% 14% 32% 10% Discontinuation before W48, n (%)16 (13%)18 (14%) For adverse event66 For virologic failure11 Baseline characteristics and patient disposition Mallolas J, JAIDS 2009;51:29-36 ATAZIP ATAZIP Study: Switch LPV/r to ATV/r

4.  Time to treatment failure and time to virological failure did not differ between groups  The median changes in CD4 count at 48 weeks were +27 cells/mm 3 (IQR: -42 to 119) with ATV/r and +48 cells/mm 3 (IQR: -5 to 112) with LPV/r (p = 0.315) Virologic rebound Treatment failure Difference estimate (95% CI) % Mallolas J, JAIDS 2009;51:29-36 ATAZIP Results: W48 outcome Switch to ATV/r Continue on LPV/r 5 7 p < 0.0001 6/1219/127 -2.1% (-8.7%, 4.2%) 0 5 10 15 20 25 30 17 20 0 5 10 15 20 25 30 p = 0.0018 21/12125/127 % -2.3% (-12.0%, 8.0%)

5. ATAZIP Study: Switch LPV/r to ATV/r 0 p = 0.043 1 2 3 -2 -3 TC/HDL-C ratio Total cholesterol/ HDL cholesterol Fasting plasma lipids changes from baseline to week 48 TriglyceridesTotal cholesterol LDL cholesterol HDL cholesterol 0 200 100 -100 -200 -300 p < 0.001 Median change from baseline (mg/dL) p < 0.001p = 0.149p = 0.185 Mallolas J, JAIDS 2009;51:29-36 ATAZIP Switch to ATV/r 300/100 qd (N = 121) Continue on LPV/r 400/100 bid (N = 127)

6. ATAZIP Study: Switch LPV/r to ATV/r ATV/r, N = 121LPV/r, N = 127 Death11 (hepatic encephalopathy) AE leading to discontinuation5% Patients with ≥ 1 Grade 3 or 4 AE11%16% Clinical CardiovascularN = 1 DigestiveN = 2N =1 OsteomuscularN = 0N = 1 HematologicN = 0N = 1 HepaticN = 1 RespiratoryN = 1N = 0 Laboratory TransaminasesN = 6N = 2 Total cholesterolN = 4N = 7 TriglyceridesN = 0N = 13 Bilirubin > 2.5 mg/dLN = 66 (55%)N = 6 (5%) Bilirubin > 5 mg/dLN = 21 (17%)N = 2 (2%) Adverse events by W48 Mallolas J, JAIDS 2009;51:29-36 ATAZIP

7. ATAZIP Study: Switch LPV/r to ATV/r  Conclusions –Switching to a simplified PI-based regimen containing ATV/r provides virological suppression and treatment failure similar to those observed with continued unmodified therapy with LPV/r –Safety and tolerability profile were similar in both groups –Improved lipid parameters were observed in the ATV/r arm –High incidence of hyperbilirubinemia occurred in the ATV/r arm –Switching patients with virologic suppression on LPV/r to once-daily ATV/r can provide an effective and well-tolerated treatment option Mallolas J, JAIDS 2009;51:29-36 ATAZIP